Ultra simple noise source does not work

I have built this noise source and tried different zeners. I have used 8.2V battery.
I cannot notice any noise in the scope. Why it does not work?
What kind of noise power should I expect?

The noise generated is to low to be seen on an oscilloscope. You have to use a spectrum analyzer.

Try this circuit, it will produce several volts of peak to peak wide band noise output.
Any available small NPN transistors should work quite well.

Its the reverse base emitter junction of the first transistor that creates all the noise, the collector is left unconnected.

This will light up your whole oscilloscope screen a nice even green glow at any reasonably fast sweep speed, it amazing.

neazoi said:

I have built this noise source and tried different zeners. I have used 8.2V battery.
I cannot notice any noise in the scope. Why it does not work?
What kind of noise power should I expect?

Click to expand...

Think about how Zener's work near the threshold with low current.
If you AC couple and use x1 probe you should see a few mV pp of thermal noise easily or more, but not 100mV on 4.3V zener

Warpspeed said:

Try this circuit, it will produce several volts of peak to peak wide band noise output.
Any available small NPN transistors should work quite well.

Its the reverse base emitter junction of the first transistor that creates all the noise, the collector is left unconnected.

This will light up your whole oscilloscope screen a nice even green glow at any reasonably fast sweep speed, it amazing.

View attachment 130799

Click to expand...

Will the useable noise cover 1-30MHz?
Will it operate ok at 9V or 8.2V as it is, or do I need to make any changes to the circuit?

Hey the 2n2904 are PNP transistors, not NPN as shown in the schematic?

Will the useable noise cover 1-30MHz?

Click to expand...

Hardly 30 MHz with the given circuit dimensioning. Presume you know a little bit about RF amplifiers and see why the bandwidth is limited.

Will it operate ok at 9V or 8.2V as it is, or do I need to make any changes to the circuit?

Click to expand...

Small signal BJT BE breakdown voltage can be expected around 7 V, means the circuit needs 9 - 10V supply voltage.

I really have no idea what the possible full bandwidth of that circuit is, but any noise generator will have a falling response with rising frequency.

There was enough output to easily toggle a high speed TTL digital counter to many Mhz for a random number generator I built a very long time ago.
The 2N2904 was what I pencilled in on my circuit at that time, sorry if that is not right.
It should work with just about any small signal fairly high frequency transistor.

The whole 12v does not appear across the reverse base emitter junction, and as I remember it worked fine on 12v and should also work at lower voltages.

This thread reminded me of my old project, and I just dipped into my archive files.

Its a great high amplitude random noise generator circuit, but it may not be ideal for every purpose.
It impressed me greatly at the time for such a simple circuit which is why I kept it for future reference.

Have a look to:

https://www.maximintegrated.com/en/app-notes/index.mvp/id/3469

Warpspeed said:

It impressed me greatly at the time for such a simple circuit which is why I kept it for future reference.

Click to expand...

Right, I know exactly what you mean, I do the same. Simple but relatively good circuits that "always work" are kept "secretly" on my website :thumbsup:

I'll give it a shot, but I would really like to see how it works on a single 9V battery. Maybe the rechargable 8.2v ones are too low, so I will try 9v alkaline ones.

You may be able to cheat with some lower value resistors and slight inductive peaking at 30 Mhz.
The basic idea is sound, if you are after more bandwidth and less amplitude a small tweak here and there might do it.

What I used it for was to drive a divide by sixteen counter, and it certainly did that at a furious multi Mhz rate.
Then I just latched the count at various times to get some truly random numbers four bits at a time.

Hey the 2n2904 are PNP transistors, not NPN as shown in the schematic?

Click to expand...

Just had a look through my stock of transistors here, trying to solve the mystery.

I do have some 2N3904's which are high frequency NPN types, and fairly likely what I decided to use. Not that I expect it is the least bit critical.

My hand writing is not the best, and I am also a bit dyslectic.
Its highly probable somewhere the 3 morphed into a 2.